A growing number of children are being implanted with magnetically controlled growing rods. Some also have baclofen pumps, a device that delivers a neuromodulating medication using a magnet. There has been no previous testing to determine if the External Remote Controller (ERC) would stall the baclofen pump during magnetic rod lengthening, a potentially life-threatening event. Two baclofen pumps, 20 mL and 40 mL, were placed in varying proximity to the ERC while the magnetic field was generated. The speed (dose) of the baclofen pump, the distance from the ERC, the length of time the ERC was run, the position of the pump in relationship to the magnetic field, and the presence or absence of a stall were recorded. To stall a baclofen pump, the ERC needed to be running continuously for at least 30 mm of lengthening (about 3 minutes). The pump would not stall if it was more than 1 cm away from the ERC, even if the ERC generated a field for 3 minutes. The speed of the pump did not affect the instance of a motor stall. The pump was more likely to stall if it was positioned in front of or behind the magnetic field, as opposed to the center of the field. Baclofen pumps are always more than 1 cm away from the magnetic rods when both are implanted in a child. The typical lengthening interval is under 10 mm, much less than what was required to stall a baclofen pump. The ERC is highly unlikely to stall a baclofen pump during a magnetic rod lengthening, making a life-threatening event almost impossible

Scoliosis is estimated to occur between 21–64% of patients with cerebral palsy (CP), where a subset of patients develops early onset scoliosis (EOS) before the age of ten. Traditional growth friendly (TGF) surgeries in the context of traditional growing rods have been shown to be effective in treating scoliosis in this population, however significant complication rates are reported. Currently, no studies have been done to examine the effect of novel growth friendly surgeries such as magnetically controlled growing rods (MCGR) on EOS in CP patients. The objective of this study is to compare MCGR with TGF surgeries in this patient population, specifically by evaluating radiographic measurements and risk of unplanned reoperations (UPRORs). Patients with EOS secondary to CP were prospectively identified from an international database, with data retrospectively analyzed. Scoliosis (primary curve), maximum kyphosis, T1-S1 and T1-T12 height were measured pre-operation, immediate post-operation, and at two-years follow-up. The risk and etiology of UPRORs were compared between MCGR and TGF. P < 0.05 was considered statistically significant for all analyses. Of the 120 patients that met inclusion criteria, 86 received TGF (age 7.5 ± 1.8 years; follow-up 7.0 ± 2.9 years) and 34 received MCGR (age 7.1 ± 2.2 years, follow-up 2.8 ± 0.5 years). Compared to TGF, MCGR resulted in significant improvements in maintenance of scoliosis correction (p=0.04). At final follow-up, UPRORs were 24% for MCGR (8/34 patients) and 43% (37/86 patients) for TGF (p=0.05). To minimize the influence of follow-up period, UPRORs within the first two years post-operation were evaluated: MCGR (21%, 7/34 patients) vs. TGF (14%, 12/86 patients; p=0.37). Within the first two years, etiology of UPROR as a percentage of all patients per group were deep infection (5% TGF, 6% MCGR), implant failure/migration (5% TGF, 9% MCGR), dehiscence (2% TGF, 3% MCGR), and superficial infection (1% TGF, 3% MCGR). The most common etiology of UPROR for TGF was deep infection and implant failure/migration and for MCGR was implant failure/migration. For patients with CP, at final follow-up, MCGR had superior maintenance of scoliosis correction; however, there was no difference in risk of UPROR within the first two years post-operatively (21% MCGR, 14% TGF)